Obviously, there are individuals and teams who do this the right way. But what about the majority of us? Will better prompts solve this? NO, a big NO. Stale documents are an industry-wide problem; anyone saying otherwise is lying.
I always ask this question: if someone in the team or company already solved a problem, why is someone else solving it again? I'm not saying a genuine attempt isn't required, I'm only talking about duplicate efforts. With AI tools, this duplication is becoming unavoidable.
So, for the last 7 months, I've been making a genuine attempt to tackle this, and I want people out here to share your thoughts: do you believe this is the right problem I'm solving? Or is there a different angle?
It is self-hosted; you can run it with Ollama. Give it a try. If you like the idea, do star the repo so I get the right audience
https://news.ycombinator.com/from?site=github.com/docbrain-a...
https://github.com/docbrain-ai/docbrain
I spend a good amount of time chasing the LLM for evidence that its answer is actually accurate. To be honest, I can usually get there, but it works great for fresh work and gets messy on months-old work. And it's not just me; a lot of people have raised the same thing.So for the last 6 months I've been trying to fix it. Today I can say it's working for me — it flags me in advance. Obviously...
https://github.com/docbrain-ai/docbrain
I build this, it’s been 6months and now finally getting real genuine people looking at it.I need real feedback, try it let me know